By Crypto Loop · Updated 2026-10-06T20:48:47.699Z
Why volatility is only a partial measure of risk
Volatility measures the dispersion of returns around an average, usually expressed as a standard deviation. That is useful, but it is only one dimension of risk. It treats upward and downward moves symmetrically, it compresses the path of returns into a single number, and it says little about how quickly losses can accumulate or whether they can be recovered before forced selling occurs.
For pillar trading, the limitation is practical. A strategy can have low day-to-day volatility and still be exposed to a small number of severe adverse moves, a series of losses that never look dramatic individually, or a liquidity shock that makes exit costly exactly when protection is needed. In other words, low volatility does not automatically mean low danger; it may simply mean the danger has not yet been realised in a measurable way.
A more complete view asks at least four questions. How large can the worst outcomes become? How often do losses cluster? How stable is the path between entry and exit? And can the position be resized, hedged, or unwound under stress without a material penalty? Volatility contributes to those answers, but it does not finish them.
Distribution tails and jump risk
Most basic risk discussions assume returns are roughly bell-shaped, meaning extreme outcomes are rare and the tails thin out quickly. Real markets often behave differently. Tail events are outcomes far from the centre of the distribution, and jump risk refers to abrupt price changes that occur between observable intervals rather than as a smooth sequence of small moves. These events matter because they can overwhelm position sizing that appears conservative under normal conditions.
A position can be designed around a tolerable average loss and still fail if the underlying moves sharply through levels used for stops or hedges. In such cases, the realised loss is determined not by the planned exit but by the price available after the jump. That difference is especially important for leveraged or derivative exposures, where small price changes in the underlying can have amplified effects on margin and liquidation thresholds.
A practical way to think about tail risk is to ask whether the strategy is robust to rare but plausible discontinuities. Examples include sudden gaps after macro announcements, exchange-specific interruptions, or concentrated market flows. The relevant issue is not whether such jumps are frequent; it is whether one jump can be large enough to negate many ordinary gains. If a strategy depends on many small profits and is vulnerable to one outsized loss, the tail shape matters more than the average return.
Realised versus implied risk
Realised volatility describes what has actually happened over a period. Implied volatility is a market-based estimate embedded in option prices, reflecting expectations and risk premia at that moment. The two can diverge materially. A market may price in turbulence that never arrives, or it may remain complacent before an abrupt repricing. Neither outcome makes the market irrational; it simply means expectations and outcomes are different objects.
For a pillar trader, the gap between realised and implied conditions is informative. If implied measures are elevated relative to realised movement, option protection may be expensive relative to recent behaviour, but that does not mean it is overpriced in a forward-looking sense. It may simply reflect uncertainty, skewed downside risk, or the market’s memory of recent stress. Conversely, low implied volatility can be a warning sign if the position’s downside is driven by jump risk rather than gradual fluctuation.
Decision checks help here. Ask whether the position is exposed mainly to continuous noise or to discrete events. Ask whether the holding period overlaps with known catalysts that can alter the distribution. Ask whether implied levels are being used as a forecast or as a price of insurance. Confusing the two can lead to under-hedging when risk appears cheap and overconfidence when historical movement looks quiet.
Drawdown is a path risk, not just an endpoint
Drawdown measures the decline from a prior peak in portfolio value. It is useful because it captures the experience of being underwater, which is often more important operationally than the volatility series itself. A strategy with moderate average volatility can still produce deep drawdowns if losses cluster or recoveries are slow. Drawdown therefore reflects both magnitude and persistence.
Path dependence matters because investors and desks rarely react to average outcomes; they react to the journey. A temporary loss can become a permanent one if capital is cut, margin is tightened, or the trader becomes unable to maintain the original thesis. Even if the position later recovers, the strategy may not survive long enough to benefit. This is one reason why risk management should not focus only on expected variance but also on the depth and duration of adverse moves.
Worked example: suppose a hypothetical position starts at 100 units of portfolio value. It falls to 92, then to 88, and later recovers to 94. The peak-to-trough drawdown is 12 units, or 12%. The return from 88 back to 100 would require a gain of about 13.6%, not 12%, because recovery percentages are measured on a smaller base. If the position is leveraged, the practical burden is greater because the required recovery can coincide with higher financing costs, tighter margin, or reduced willingness to hold through the drawdown. This is why a position that appears only mildly volatile can still be hard to own.
Liquidity risk: the cost of getting out
Liquidity risk is the risk that a position cannot be transacted quickly, in size, and at a predictable cost. It is often hidden in calm markets because quotes appear available and spreads look tight. Under stress, however, liquidity can thin out, spreads can widen, and order book depth can disappear. The consequence is slippage: the difference between the expected execution price and the actual fill price.
For pillar trading, liquidity matters in two ways. First, the asset itself may be less liquid than the benchmark used to judge risk. Second, even if the asset is liquid in normal conditions, the size of the position relative to market depth may create self-induced illiquidity. That means the act of exiting can move the market against the trader. The risk is not only losing money on paper; it is losing the ability to control the loss.
Practical checks are simple but important. Compare intended position size with normal turnover rather than with headline market capitalisation. Examine whether the trade can be reduced in increments without crossing a large spread. Consider whether hedges are equally liquid or if the hedge itself would be difficult to unwind. And ask what happens if liquidity is available only at worse prices for a limited window. In many cases, liquidity risk is the mechanism that turns a manageable price move into an unacceptable realised loss.
Operational risk and implementation failure
Operational risk covers errors and failures in systems, people, processes, and counterparties. In trading, this includes stale data, mispriced instruments, missed alerts, incorrect sizing, failed order routing, wrong contract expiry, incomplete reconciliation, and custody or settlement problems. These issues are not market risk in the narrow sense, but they often determine whether market risk is contained or amplified.
Operational failures matter because pillar trading often relies on rules. A strategy may define stop levels, rebalancing thresholds, hedge ratios, or expiry dates, but those rules only protect capital if they are implemented correctly and on time. A missed hedge during a fast move, an unnoticed duplicate order, or an error in contract selection can create exposures that were never intended. The failure may be small at first and still become large through compounding.
A useful control question is whether the strategy still behaves acceptably if one component fails. For example, if a hedge cannot be entered, can the core position survive until manual review? If market data is delayed, does the system freeze trading or continue blindly? If a counterparty becomes unavailable, is there a documented fallback? Operational resilience is part of risk, not a separate administrative detail. It determines whether the theoretical risk model matches the realised outcome.
A practical framework for deciding whether risk is acceptable
A useful framework begins with three layers. Layer one is distribution shape: how much of the risk sits in the tails rather than in the centre. Layer two is path and liquidity: how losses would unfold and whether they can be exited or hedged without severe penalty. Layer three is implementation: whether the plan can actually be executed under stress. A position that looks acceptable on only one layer can still fail on the others.
One practical sequence is to stress the trade with questions rather than forecasts. What if the market gaps through the planned stop? What if implied conditions reprice before the catalyst? What if the hedge is less liquid than the exposure? What if the position must be reduced while spreads are wide? What if execution is delayed by a platform or data issue? These are not predictions; they are tests of resilience.
Another useful check is to distinguish tolerable volatility from intolerable loss. Some strategies can live with frequent small fluctuations but not with deep drawdowns. Others can accept mark-to-market noise if the liquidity and tail profile are favourable. The right fit depends on capital constraints, time horizon, and the consequences of being wrong. Risk is acceptable only if the trader can survive the bad path, not just the average one.
Failure scenarios and the limits of standard metrics
Standard metrics fail most obviously when returns are non-linear. A calm volatility history may hide exposure to sudden jumps, short-option-like convexity, or regime shifts where correlations change at the worst moment. Metrics based on averages can also understate concentration risk if several positions depend on the same underlying driver, because the measured volatility of each line item may look modest while the combined portfolio is fragile.
A second failure scenario appears when liquidity vanishes. In a stress event, correlations often rise, spreads widen, and exit prices deteriorate at the same time. A portfolio that was diversified in normal conditions can become crowded into the same side of the market. If risk controls assume continuous trading, they can break exactly when they are most needed.
The limits of volatility are therefore structural, not technical. Volatility ignores tail thickness, says little about skew, and does not measure the cost of crossing the spread, funding a position, or making an operational mistake. It is still worth monitoring, but only as one input among several. If a trader uses volatility as the whole of risk, they may be surprised not by ordinary noise but by the first event that breaks the assumptions behind the metric.
Jurisdiction and risk caveat
This article is educational and general in nature, not legal, tax, accounting, or investment advice. Rules on trading, leverage, derivatives, reporting, margin, client classification, and market access vary by jurisdiction and can change over time. Any reader should confirm the relevant obligations and product restrictions in their own location before trading or designing a risk process.
More broadly, the examples here are hypothetical and simplified. Real markets involve instrument-specific terms, venue rules, financing conditions, and operational constraints that can materially change outcomes. A conservative approach is to treat volatility as an early warning indicator, not as a complete risk measure, and to validate any strategy against tails, drawdowns, liquidity, and execution failure before committing capital.